• Ahmed Abaza

Is Artificial Intelligence, Intelligent?

Updated: Jun 9

An article about the I in AI

To answer this question, one might need to define what is intelligence to begin with? This is a question, that is so controversial, it has so many opinions, researchers and thought leaders are still trying to frame it in the most accurate way. I am not going to define intelligence here, me being the last one on earth to do so (Not for the lack thereof, but, Well, I disappoint myself sometimes)

Though, for the sake of context I will categorize intelligence in two main categories. The first one being, Abstract Intelligence. This is where, us humans are able to conceptualize, theorize and imagine. This is your philosophical self .Take gravity for example, humans knew about gravity, and conceptualized it in different formats since the beginning of time. Newton came and theorized a concept for gravity, that is a force must be acting on falling objects because otherwise they would not start moving from rest and has put this concept for this force (F) into an equation that goes: F = Gm1m2/r2, by legend has it, observing an apple falling on his head. Newton here showcased a display of the other type of intelligence that is the “Concrete Intelligence”. It is basically observing a certain pattern, and describing this phenomena into an equation. This equation can be leveraged to model the world, and create calculability , by which humans were designing such things like Airplanes.

This is the type of intelligence, in which, Artificial Intelligence shines, the "Concrete Intelligence"

You feed an Artificial Intelligence algorithm a lot of data (Observations), collected from a certain phenomena (Think lots of apples falling down), and it puts it into an equation. This equation can then be used to automate, calculate or predict things.

Yup Newton Was Right!

For example, you can show an A.I a big enough data set for your sales data, and it creates a model that helps you to predict your sales more accurately, or it can fuel a recommendation engines, that help you sell more to your customers.

Another example, that Artificial Intelligence can kick humans bottoms in chess (Concrete Intelligence), but it cannot conceptualise the chess board (Abstract Intelligence) by itself. This is where I think the business hype about artificial intelligence might be a little misled.

Maybe it was Einstein who once said, that we cannot solve a problem with the same thinking we used when we created them. A.I is amazing, at making our frames of thinking much more calculable and efficient, though, it will still not be transformative enough, if it’s restrained to our preset mind frames.

I am not saying that Artificial Intelligence with it's current state is not effective, on the contrary. I witnessed it first hand, through helping clients with my Company, Synapse Analytics, achieve some amazing optimisations, cost savings and growth, that wouldn't have been possible without A.I. Any business in this day and age demands calculability that gives control, and predictability that empowers smarter decision making. This is the foundation for any sustainable and competitive modern business.

What I am arguing is that artificial intelligence within the current modality of thinking within the business world, is like wanting to go to the moon by building the empire state, you will be way taller than any other building, it's an amazing achievement, and everybody will be shorter than you, though, you will not reach the moon.

This also sheds light on the kind of talent you should have within your data science and A.I teams. It's easy to code algorithms, and A.I is now very accessible as a technology. Though, the real work happens in the problem design and framing. This dictates how one models the data, chooses and tunes the right algorithm to make a transformative optimisation or create a powerful impact. Otherwise, one might fall in many traps such as biases or build a useless, highly accurate algorithms.

Take for example the the famous hiring algorithm case, where a recruitment algorithm was biased against women. One obvious fact why this happened, because it was trained on the same data, that was biased against women. So, it only empowered misjudgement, through observing the wrong observations. It's like having Newton watch apples falling up to the sky, he would have come up with a completely different equation.

Is it possible for A.I to reach Abstraction?

Yes, or at least I believe so, maybe within the next few decades. Is this good or bad? That’s for another post. There is so much research going towards that and progress already, within the General A.I, self aware and super intelligence. I will be mentioning their progress and achievement in a future post.

Though, for now, for current A.I use to be powerful and transformative, the technology alone won’t do it. It still requires us to do the abstraction work, the problem design, to frame new modalities of thinking, knowing that we have a very powerful lever to change the world, that is A.I. Or, Maybe unleash A.I to become it's own designer, and build it's own predictive variables and goals, who knows?

Founder & CEO of Synapse Analytics on how AI can be used in Retail. Synapse is a data science and AI company based in Egypt. Synapse Analytics





Digital Marketing & Advertising

Distribution & Logistics



Real Estate



Financial Services




By Use Case

Credit Scoring

Demand Forecasting

Customer Segmentation

Warehouse Location Optimization

Copyright © 2020 Synapse Analytics, All rights reserved.

  • Facebook
  • LinkedIn
  • YouTube
  • Twitter
  • Instagram